A CCA and ICA-Based Mixture Model for Identifying Major Depression Disorder

Three layered sparse dictionary learning algorithm for enhancing the subject wise segregation of brain networks

Independent component analysis (ICA) and dictionary learning (DL) are the most successful blind source separation (BSS) methods for functional magnetic resonance imaging (fMRI) data analysis. However, ICA to higher and DL to lower extent may suffer from performance degradation by the presence of anomalous observations in the recovered time courses (TCs) and high overlaps among spatial maps (SMs). This paper addressed both problems using a novel three-layered sparse DL (TLSDL) algorithm that incorporated prior information in the dictionary update process and recovered full-rank outlier-free TCs from highly corrupted measurements. The associated sequential DL model involved factorizing each subject's data into a multi-subject (MS) dictionary and MS sparse code while imposing a low-rank and a sparse matrix decomposition restriction on the dictionary matrix. It is derived by solving three layers of feature extraction and component estimation. The first and second layers captured brain regions with low and moderate spatial overlaps, respectively. The third layer that segregated regions with significant spatial overlaps solved a sequence of vector decomposition problems using the proximal alternating linearized minimization (PALM) method and solved a decomposition restriction using the alternating directions method (ALM). It learned outlier-free dynamics that integrate spatiotemporal diversities across brains and external information. It differs from existing DL methods owing to its unique optimization model, which incorporates prior knowledge, subject-wise/multi-subject representation matrices, and outlier handling. The TLSDL algorithm was compared with existing dictionary learning algorithms using experimental and synthetic fMRI datasets to verify its performance. Overall, the mean correlation value was found to be 26 % higher for the TLSDL than for the state-of-the-art subject-wise sequential DL (swsDL) technique.

Classification of MDD using a Transformer classifier with large-scale multisite resting-state fMRI data

Major depressive disorder (MDD) is one of the most common psychiatric disorders worldwide with high recurrence rate. Identifying MDD patients, particularly those with recurrent episodes with resting-state fMRI, may reveal the relationship between MDD and brain function. We proposed a Transformer-Encoder model, which utilized functional connectivity extracted from large-scale multisite rs-fMRI datasets to classify MDD and HC. The model discarded the Transformer's Decoder part, reducing the model's complexity and decreasing the number of parameters to adapt to the limited sample size and it does not require a complex feature selection process and achieves end-to-end classification. Additionally, our model is suitable for classifying data combined from multiple brain atlases and has an optional unsupervised pre-training module to acquire optimal initial parameters and speed up the training process. The model's performance was tested on a large-scale multisite dataset and identified brain regions affected by MDD using the Grad-CAM method. After conducting five-fold cross-validation, our model achieved an average classification accuracy of 68.61% on a dataset consisting of 1611 samples. For the selected recurrent MDD dataset, the model reached an average classification accuracy of 78.11%. Abnormalities were detected in the frontal gyri and cerebral cortex of MDD patients in both datasets. Furthermore, the identified brain regions in the recurrent MDD dataset generally exhibited a higher contribution to the model's performance.

A novel subject-wise dictionary learning approach using multi-subject fMRI spatial and temporal components

The conventional dictionary learning (DL) algorithms aim to adapt the dictionary/sparse code to individual functional magnetic resonance imaging (fMRI) data. Thus, lacking the capability to consolidate the spatiotemporal diversities offered by other subjects. Considering that subject-wise (sw) data matrix can be decomposed into the sparse linear combination of multi-subject (MS) time courses and MS spatial maps, two new algorithms, sw sequential DL (swsDL) and sw block DL (swbDL), have been proposed. They are based on the novel framework, defined by the mixing model, where base matrices prepared by operating a computationally fast sparse spatiotemporal blind source separation method over multiple subjects are employed to adapt the mixing matrices to sw training data. They solve the optimization models formulated using [Formula: see text]/[Formula: see text]-norm penalization/constraints through dictionary/sparse code pair update and alternating minimization approach. They are unique because no existing sparse DL method can incorporate MS spatiotemporal components while updating sw atoms/sparse codes, which can eventually be assembled using neuroscience knowledge to extract group-level dynamics. Various fMRI datasets are used to evaluate and compare the performance of the proposed algorithms with existing state-of-the-art algorithms. Specifically, overall, a [Formula: see text] increase in the mean correlation value and [Formula: see text] reduction in the mean computation time exhibited by swsDL and swbDL, respectively, over the adaptive consistent sequential dictionary algorithm.

Intra and inter: Alterations in functional brain resting-state networks in patients with functional constipation

Background

Functional constipation (FCon), is a symptom-based functional gastrointestinal disorder without an organic etiology and altering brain structure and function. However, previous studies mainly focused on isolated brain regions involved in brain plasticity. Therefore, little is known about the altered large-scale interaction of brain networks in FCon.

Methods

For this study, we recruited 20 patients with FCon and 20 healthy controls. We used group independent component analysis to identify resting-state networks (RSNs) and documented intra- and inter-network alterations in the RSNs of the patients with FCon.

Results

We found 14 independent RSNs. Differences in the intra-networks included decreased activities in the bilateral caudate of RSN 3 (strongly related to emotional and autonomic processes) and decreased activities in the left precuneus of RSN 10 (default mode network). Notably, the patients with FCon exhibited significantly decreased interactive connectivity between RSNs, mostly involving the connections to the visual perception network (RSN 7–9).

Conclusion

Compared with healthy controls, patients with FCon had extensive brain plastic changes within and across related RSNs. Furthermore, the macroscopic brain alterations in FCon were associated with interoceptive abilities, emotion processing, and sensorimotor control. These insights could therefore lead to the development of new treatment strategies for FCon.

Intra and inter: Alterations in functional brain resting-state networks after peripheral nerve injury

INTRODUCTION

Numerous treatments suggest that brain plasticity changes after peripheral nerve injury (PNI), and most studies examining functional magnetic resonance imaging focused on abnormal changes in specific brain regions. However, it is the large-scale interaction of neuronal networks instead of isolated brain regions contributed to the functional recovery after PNI. In the present study, we examined the intra- and internetworks alterations between the related functional resting-state networks (RSNs) in a sciatic nerve injury rat model.

METHODS

Ninety-six female rats were divided into a control and model group. Unilateral sciatic nerve transection and direct anastomosis were performed in the latter group. We used an independent component analysis (ICA) algorithm to observe the changes in RSNs and assessed functional connectivity between different networks using the functional networks connectivity (FNC) toolbox.

RESULTS

Six RSNs related to PNI were identified, including the basal ganglia network (BGN), sensorimotor network (SMN), salience network (SN), interoceptive network (IN), cerebellar network (CN), and default mode network (DMN). The model group showed significant changes in whole-brain FC changes within these resting-state networks (RSNs), but four of these RSNs exhibited a conspicuous decrease. The interalterations performed that significantly decreased FNC existed between the BGN and SMN, BGN and IN, and BGN and DMN (p < .05, corrected). A significant increase in FNC existed between DMN and CN and between CN and SN (p < .05, corrected).

CONCLUSION

The results showed the large-scale functional reorganization at the network level after PNI. This evidence reveals new implications to the pathophysiological mechanisms in brain plasticity of PNI.

Major depressive disorder identification by referenced multiset canonical correlation analysis with clinical scores

A novel method based on multiset canonical correlation analysis (mCCA) and linear discriminant analysis (LDA) is presented to identify the major depressive disorder (MDD). The new method comprises two parts, namely, the mCCA-rreg and sparse LDA models. The mCCA-rreg model extends the classical canonical correlation model to calculate functional connections by restricting the references to a reference space and adding a spatial regularization term. The reference space is used to ensure that the model extracts important components first from several datasets simultaneously by decreasing the importance of the components in which we are uninterested. The spatial regularization term helps in avoiding the multicollinearity and overfitting problems under the low signal-to-noise ratio circumstance. The sparse LDA model extends the classical LDA model to extract a small subset of discriminative classification features by fusing clinical scores. In the real data experiment, we extract two functional connection modes from 45 subjects by the mCCA-rreg model. Then, we construct classifiers to identify the patients with MDD based on the connections selected by the sparse LDA model. The best accuracy is higher than 95%. The results show that the mCCA-rreg model can retrieve the important components characterized by a preassigned reference space and exclude the noise or components of no interest. The sparse LDA model can extract discriminative classification features related to clinical scores.

A Projection CCA Method for Effective fMRI Data Analysis

OBJECTIVE

Canonical correlation analysis (CCA) is a data-driven method that has been successfully used in functional magnetic resonance imaging (fMRI) data analysis. Standard CCA extracts meaningful information from a pair of data sets by seeking pairs of linear combinations from two sets of variables with maximum pairwise correlation. So far, however, this method has been used without incorporating prior information available for fMRI data. In this paper, we address this issue by proposing a new CCA method named pCCA (for projection CCA).

METHODS

The proposed method is obtained by projection onto a set of basis vectors that better characterize temporal information in the fMRI data set. A methodology is presented to describe the basis selection process using discrete cosine transform (DCT) basis functions. Employing DCT guides the estimated canonical variates, yielding a more computationally efficient CCA procedure.

RESULTS

The performance gain of the proposed pCCA algorithm over standard and regularized CCA is illustrated on both simulated and real fMRI datasets from resting state, block paradigm task-related and event-related experiments. The results have shown that the proposed pCCA successfully extracts latent components from the task as well as resting-state datasets with increased specificity of the activated voxels.

CONCLUSION

In addition to offering a new CCA approach, when applied on fMRI data, the proposed algorithm adapts to variations of brain activity patterns and reveals the functionally connected brain regions.

SIGNIFICANCE

The proposed method can be seen as a regularized CCA method where regularization is introduced via basis expansion, which has the advantage of enforcing smoothness on canonical components.

The development of brain functional connectivity networks revealed by resting-state functional magnetic resonance imaging

Previous studies on brain functional connectivity networks in children have mainly focused on changes in function in specific brain regions, as opposed to whole brain connectivity in healthy children. By analyzing the independent components of activation and network connectivity between brain regions, we examined brain activity status and development trends in children aged 3 and 5 years. These data could provide a reference for brain function rehabilitation in children with illness or abnormal function. We acquired functional magnetic resonance images from 15 3-year-old children and 15 5-year-old children under natural sleep conditions. The participants were recruited from five kindergartens in the Nanshan District of Shenzhen City, China. The parents of the participants signed an informed consent form with the premise that they had been fully informed regarding the experimental protocol. We used masked independent component analysis and BrainNet Viewer software to explore the independent components of the brain and correlation connections between brain regions. We identified seven independent components in the two groups of children, including the executive control network, the dorsal attention network, the default mode network, the left frontoparietal network, the right frontoparietal network, the salience network, and the motor network. In the default mode network, the posterior cingulate cortex, medial frontal gyrus, and inferior parietal lobule were activated in both 3- and 5-year-old children, supporting the “three-brain region theory” of the default mode network. In the frontoparietal network, the frontal and parietal gyri were activated in the two groups of children, and functional connectivity was strengthened in 5-year-olds compared with 3-year-olds, although the nodes and network connections were not yet mature. The high-correlation network connections in the default mode networks and dorsal attention networks had been significantly strengthened in 5-year-olds vs. 3-year-olds. Further, the salience network in the 3-year-old children included an activated insula/inferior frontal gyrus-anterior cingulate cortex network circuit and an activated thalamus-parahippocampal-posterior cingulate cortex-subcortical regions network circuit. By the age of 5 years, nodes and high-correlation network connections (edges) were reduced in the salience network. Overall, activation of the dorsal attention network, default mode network, left frontoparietal network, and right frontoparietal network increased (the volume of activation increased, the signals strengthened, and the high-correlation connections increased and strengthened) in 5-year-olds compared with 3-year-olds, but activation in some brain nodes weakened or disappeared in the salience network, and the network connections (edges) were reduced. Between the ages of 3 and 5 years, we observed a tendency for function in some brain regions to be strengthened and for the generalization of activation to be reduced, indicating that specialization begins to develop at this time. The study protocol was approved by the local ethics committee of the Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences in China with approval No. SIAT-IRB-131115-H0075 on November 15, 2013.